There are three Yersinia species that are pathogenic to humans: Y. pestis, Y. pseudotuberculosis, and Y. enterocolitica. Y. pestis is the causative agent of plague, while Y. pseudotuberculosis and specific pathogenic serovars of Y. enterocolitica cause gastrointestinal illnesses. Other species of Yersinia, including Y. rohdei, Y. aldovae, Y. bercovieri, Y. frederiksenii, Y. intermedia, Y. kristensenii, and Y. moolaretti, are considered enterocolitica-like opportunist pathogens with the ability to cause diarrheal illness in susceptible individuals (Agbonlahor, J Clin Microbiol, 23, 891-6, (1986), Cafferkey, et al., J Hosp Infect, 24, 109-15, (1993), Loftus, et al., Dig Dis Sci, 47, 2805-10, (2002)). The Yersinia can also infect other animal species causing a range of illnesses. Most wild and domestic species of mammals are prone to infections with the enteropathogens Y. enterocolitica and Y. pseudotuberculosis, although most of these infections are subclinical and such animals usually serve only as asymptomatic carriers of the pathogens for transmission to humans (Fantasia, et al., J Clin Microbiol, 22, 314-5, (1985), Fantasia, et al., Vet Rec, 132, 532-4, (1993), Fukushima, et al., J Clin Microbiol, 18, 981-2, (1983), Kageyama, et al., J Med Primatol, 31, 129-35, (2002), Kato, et al., Appl Environ Microbiol, 49, 198-200, (1985), Poelma, et al., Acta Zool Pathol Antverp, 3-9, (1977), Shayegani, et al., Appl Environ Microbiol, 52, 420-4, (1986), Yanagawa, et al., Microbiol Immunol, 22, 643-6, (1978).). However, there are reports that the enteropathogenic Yersinia have been associated with diarrheal illness and general malaise in domestic animals such as sheep, cattle, goats, pigs, dogs, birds and farmed deer (Jerrett, I. V., et al., Aust Vet J, 67, 212-4, (1990), Slee, K. J., et al., Aust Vet J, 65, 271-5, (1988), Slee, K. J. and C. Button, Aust Vet J, 67, 396-8, (1990), Slee, K. J. and C. Button, Aust Vet J, 67, 320-2, (1990), Zheng, X. B., J Appl Bacteriol, 62, 521-5, (1987)). Y. pestis can cause disease in a variety of rodent species as well as nonhuman primates (Davis, K. J., et al., Arch Pathol Lab Med, 120, 156-63, (1996), Meyer, K. F., et al., J Infect Dis, 129, Suppl:S85-12, (1974). Y. pestis is also associated with potentially severe infections in domestic cats (Gasper, P. W., et al., J Med Entomol, 30, 20-6, (1993)) and a few cases of Y. pestis infection have been reported in dogs (Orloski, K. A. and M. Eidson, J Am Vet Med Assoc, 207, 316-8, (1995)). In addition, Yersinia ruckeri is a pathogen of fish, causing redmouth disease in salmonids (Furones, M. D., et al., Ann. Rev. Fish Dis., 3, 105-125, (1993)).
Plague is undoubtedly one of the most devastating acute infectious disease in the recorded history of man, estimated to have killed 100 to 200 million people worldwide (Perry, R. D. and J. D. Fetherston, Clin Microbiol Rev, 10, 35-66, (1997)). In recent years plague outbreaks have been relatively uncommon in the U.S. and other industrialized countries, although endemic foci exist in all continents except Australia. Worldwide surveys indicated 2000 to 5000 annual cases of plague reported in the last several years, although epidemiologists suspect that many human cases of plague are unreported. Y. pseudotuberculosis outbreaks are fairly rare, and have occurred primarily in Finland, Japan, and the former Soviet Union (Inoue, M., et al., Zentralbl Bakteriol Mikrobiol Hyg [B], 186, 504-511, (1988), Nuorti, J. P., et al., J Infect Dis, 189, 766-774, (2004), Rodina, L. V., et al., Zh Mikrobiol Epidemiol Immunobiol, 116-118, (1998), Toyokawa, Y., et al., Kansenshogaku Zasshi, 67, 36-44, (1993)). Most Y. pseudotuberculosis infections are assumed to be transmitted by the oral-fecal route; however, a vehicle of transmission has not been identified in many cases. In the United States, infections by Y. enterocolitica are more common than those with Y. pseudotuberculosis, and are typically associated with the consumption of contaminated pork products (Ray, S. M., et al., Clin Infect Dis, 38 Suppl 3, S181-189, (2004)). The incidence of human disease caused by the Y. enterocolitica in the U.S. is difficult to determine, simply because infections associated with this organism are typically self-limiting and insufficient detection techniques have limited the ability to correctly diagnose the causative agent. However, FoodNet surveillance for 1996-1999 estimated approximately 1 case of Y. enterocolitica infection per 100,000 in the United States (Ray, S. M., et al., Clin Infect Dis, 38 Suppl 3, S181-189, (2004)).
Plague is an infectious disease of animals and humans having both enzootic and epizootic components of transmission. The most naturally occurring means of transmission is from an infected rodent reservoir to fleas, which serve as natural vectors for transmission to humans. However, human-to-human transmission can also occur by direct contact or respiratory inhalation of contaminated droplets (Pneumonic form). Nevertheless, in natural infections Y. pestis typically enter humans by a subcutaneous route into the bloodstream, where they travel to the lymph nodes and begin to multiply. Clinical manifestations of plague include large swollen masses near the lymph nodes, referred to as bubos. Occasionally, Y. pestis multiplies rapidly in the bloodstream, inducing septicemia with an accompanying general malaise that includes fever, headache, chills, and occasionally gastrointestinal disturbances. These symptoms are often misdiagnosed early, and antibiotic therapy may therefore be administered too late for effective intervention. Septicemic infection by Y. pestis has a 50% fatality rate (Perry, R. D. and J. D. Fetherston, Clin Microbiol Rev, 10, 35-66, (1997)), and can lead to pulmonary infection. The pneumonic form of plague is extremely infectious by the aerosol route and is characterized by a rapid onset of disease and a mortality rate close to 100%. Therefore, although antibiotic therapies are available and effective if administered early, the rapid onset of pneumonic plague and the misdiagnosis of septicemic plague are major obstacles in treatment of the disease.
Y. enterocolitica and Y. pseudotuberculosis are considered enteropathogens since most human infections are transmitted by the fecal-oral route and are limited to the gastrointestinal tract. In a normal host, Y. enterocolitica causes a diarrheal illness, which may be accompanied by fever and lower quadrant pain that mimics appendicitis. Y. pseudotuberculosis typically does not cause diarrheal illness, and is more likely to cause mesenteric lymphadenitis which can be misdiagnosed as appendicitis. Following ingestion, both organisms attach to the intestinal lymphoid tissues and traverse the mucosal layer, where they can subsequently multiply in the mesenteric lymph nodes and migrate to the spleen and liver (Lian, C. J., et al., J Med Microbiol, 24, 219-226, (1987), Une, T., Microbiol Immunol, 21, 505-516, (1977)). Y. pseudotuberculosis and some serotypes of Y. enterocolitica can also spread to the vascular system and cause fatal cases of septicemia (Bottone, E. J., Clin. Microbiol. Rev., 10, 257-276, (1997), Lenz, T., et al., J Infect Dis, 150, 963, (1984)), although these more invasive infections are typically limited to susceptible individuals. Y. enterocolitica has also been associated with septicemia following blood transfusions; in these cases, the blood supply was contaminated with the organism, which can survive and grow at refrigeration temperatures (Natkin, J. B., K G, Clin Lab Med, 19, 523-536, (1999)). Furthermore, intestinal Yersinia infections can lead to delayed sequelae such as reactive arthritis and thyroiditis (Bottone, E. J., Clin. Microbiol. Rev., 10, 257-276, (1997), Gaston, J. S., et al., Arthritis Rheum., 42, 2239-2242, (1999), Taccetti, G., et al., Clin Exp Rheumatol, 12, 681-684, (1994)). Antibiotic therapy has not been demonstrated to reduce the severity or duration of gastrointestinal illness caused by these two pathogens (Hoogkamp-Korstanje, J., J Antimicrob Chemother, 20, 123, (1987), Pai, C. H., et al., J Pediatr, 104, 308-11, (1984)). However, a susceptible host is typically treated with antibiotics to prevent more serious clinical manifestations of disease. Septicemia caused by either of these enteropathogens is also generally treated with antibiotics, and such therapies are frequently successful against Y. enterocolitica (Gayraud, M., et al., Clin Infect Dis, 17, 405-10, (1993)). In contrast, antibiotic therapy has traditionally been less effective in patients where septicemia is caused by Y. pseudotuberculosis, and the mortality rate associated with Y. pseudotuberculosis septicemia is approximately 75% (Natkin, J. B., K G, Clin Lab Med, 19, 523-536, (1999)).
Although natural infection by Y. pestis is rare in this country, there is fear that the organism will become a bioterrorism agent. As a tool of deliberate mass infection, the Y. pestis organism is a prime candidate due to several characteristics. First, the organism is highly infectious when spread by aerosol, a convenient method of mass dissemination. Second, there is a high mortality rate associated with Y. pestis infection if left untreated, and the pneumonic form of plague is distinguished by a rapid onset of symptoms that may be recognized too late for an effective intervention. Finally, Y. pestis has a well-defined genetic system, thus antibiotic-resistant strains are relatively easy to engineer.
Several plague vaccines with varying levels of efficacy and safety have been investigated. One of the earliest vaccines consisted of killed whole cells (KWC) of Y. pestis; this type of vaccine was first used in the late 1890's and confers protection against the bubonic form of plague. However, there is evidence that KWC immunizations offer little protection against pneumonic plague (Cohen, R. J. and J. L. Stockard, JAMA, 202, 365-366, (1967), Meyer, K. F., Bull World Health Organ, 42, 653-666, (1970)), and an additional drawback to these vaccines is that multiple injections over several months are required for protective immunity. An attenuated strain of Y. pestis, strain EV76, has been studied as a live vaccine for plague. In mouse studies, this vaccine has been shown to protect against both subcutaneous and inhalation challenges and requires as few as one dose for protection (Russell, P., et al., Vaccine, 13, 1551-1556, (1995)). However, strain EV76 is not fully avirulent, causing death in approximately 1% of vaccinated mice (Russell, P., et al., Vaccine, 13, 1551-1556, (1995)). Interestingly, there have been several unsuccessful attempts to create an avirulent strain of Y. pestis suitable for use as a live vaccine (Titball, R. W. and E. D. Williamson, Vaccine, 19, 4175-4184, (2001)).
Subunit vaccines are considered to be the most promising type of vaccine for safe and effective prevention of plague, primarily because there is no fear of adverse effects in a human host. Several surface proteins associated with Yersinia virulence were tested for their immunogenicity; all of these proteins induced an antibody response but only the F1 capsule and the secreted V antigen elicited good protection against challenge (Titball, R. W. and E. D. Williamson, Vaccine, 19, 4175-4184, (2001)). Both F1 and V antigen provide protection as individual antigens in animal models, although the combination of the two antigens provides superior protection. Many recent studies have tested F1/V vaccines formulated with alternative adjuvants in an attempt to find the best delivery system for the F1 and V antigens (Alpar, H. O., et al., Adv. Drug Deliv. Rev., 51, 173-201, (2001), Eyles, J. E., et al., J Control Release, 63, 191-200, (2000), Jones, S. M., et al., Vaccine, 19, 358-366, (2001), Reddin, K. M., et al., Vaccine, 16, 761-767, (1998), Williamson, E. D., et al., Vaccine, 19, 566-571, (2000), Williamson, E. D., et al., Vaccine, 14, 1613-9, (1996)).
Other innovative strategies have used attenuated Salmonella strains as vaccine carriers for Y. pestis antigens. When a Salmonella aroA mutant expressing an F1/V fusion protein was used as a vaccine strain, 86% of mice survived a subsequent lethal challenge dose of Y. pestis (Leary, S. E., et al., Microb Pathog, 23, 167-179, (1997)). Similarly, a vaccine consisting of a DNA plasmid bearing a gene encoding truncated-F1 capsule provided 80 to 100% protection in different mouse strains (Grosfeld, H., et al., Infect Immun, 71, 374-383, (2003)). In addition, a group of investigators mapped the B- and T-cell epitopes of the F1 antigen and utilized the immunoreactive peptides in vaccine formulations (Sabhnani, L., et al., FEMS Immunol Med Microbiol, 38, 215-29, (2003)). Their results indicated that a mixture of epitopic peptides protected 83% of mice against a lethal dose of Y. pestis. 
In contrast to the extensive search for protective plague vaccines, very little research efforts have been focused on preventing infections by the enteropathogenic Yersinia species. However, a few studies have demonstrated promising results. For example, attenuated Y. enterocolitica strains administered orally to mice displayed protective effects, reducing the bacterial load in the spleen and liver following oral challenge (Igwe, E. I., et al., Infect Immun, 67, 5500-5507, (1999)). However, these strains were engineered primarily as live oral vaccine carriers, and no further testing of these strains for prevention of yersiniosis has been reported. Two subunit vaccines were demonstrated as effective in animal models of infection. The first consisted of cellular extracts from Y. enterocolitica and was administered intranasally to mice. The immunized mice demonstrated enhanced clearance of an intranasal challenge dose of Y. enterocolitica from the lungs (Di Genaro, M. S., et al., Microbiol. Immunol., 42, 781-788, (1998)). A second subunit vaccine was formulated using a heat shock protein HSP60 from Y. enterocolitica adjuvanted with interleukin-12 (Noll, A. and AutenriethIb, Infect Immun, 64, 2955-2961, (1996)). Immunizations with this vaccine resulted in significantly fewer bacteria in mouse spleens following challenge, illustrating a protective effect. Additional work utilized a vaccine consisting of DNA encoding the Y. enterocolitica HSP60 in intramuscular immunizations in mice (Noll, A., et al., Eur J Immunol, 29, 986-996, (1999)). This study demonstrated that hsp60 mRNA was present in various host tissues following immunization, but protection against Y. enterocolitica challenge was limited to the spleen and no protection was observed in the intestinal mucosa.
The similarities and differences between the diseases caused by the pathogenic Yersinia species have been the focus of much research in the past decade. This is partly due to several observations that suggest the pathogenic Yersinia provide a useful model of pathogen evolution. First, DNA hybridization studies and recent genomic comparisons of fully sequenced Y. pestis and Y. pseudotuberculosis strains have indicated that these two pathogens are highly related (Chain, P. S., et al., Proc. Natl. Acad. Sci. USA, 101, 13826-13831, (2004), Ibrahim, A., et al., FEMS Microbiol Lett, 114, 173-177, (1993)), and it has been estimated that Y. pestis evolved from Y. pseudotuberculosis as recently as 1,500 to 20,000 years ago (Achtman, M., et al., Proc. Natl. Acad. Sci. USA, 96, 14043-14048, (1999)). However, despite their close evolutionary relationship, Y. pseudotuberculosis and Y. pestis cause very different diseases in humans. Furthermore, partial sequencing and 16s RNA hybridization studies suggested that Y. enterocolitica is more distantly related to the other pathogenic species of this genus (Ibrahim, A., et al., FEMS Microbiol Lett, 114, 173-177, (1993), Moore, R. L. and R. R. Brubaker, Int J Syst Bacteriol, 25, 336-339, (1975)), although Y. enterocolitica causes gastrointestinal infections similar to those observed with Y. pseudotuberculosis. Recent research has thus been focused on the virulence genes of the three pathogenic Yersinia species in an attempt to elucidate the different mechanisms they employ to cause disease. Mouse models have been particularly instructive in studying Yersinia pathogenesis, since all three species cause similar diseases in mice when injected intravenously, and more natural infections can be effectively simulated through oral and pneumonic challenge routes in mice.
A few virulence factors are unique to Y. pestis. These include proteins encoded on the Y. pestis plasmids pPCP and pMT, plasmids that are not found in Y. enterocolitica or Y. pseudotuberculosis. The pPCP plasmid encodes the plasminogen activator, a protein involved in rapid dissemination of bacteria into mammalian host tissues following subcutaneous injection (Sodeinde, O. A., et al., Science, 258, 1004-1007, (1992)). The pMT plasmid harbors at least two genes that aid in the infection of non-human hosts. The pMT-encoded caf1 gene is required for assembly of the F1 capsule, a factor that inhibits phagocytosis in the murine host but is not required for virulence in primates (Friedlander, A. M., et al., Clin. Infect. Dis., 21 Suppl 2, S178-181, (1995)). The murine toxin is also encoded on the pMT plasmid, and is believed to promote survival in the flea although it is not a required virulence factor in murine hosts (Hinnebusch, B. J., et al., Science, 296, 733-735, (2002), Hinnebusch, J., et al., Int J Med Microbiol, 290, 483-487, (2000)). Other differences between the species are the structures of the lipopolysaccharide (LPS) molecules produced by the yersiniae. Both Y. enterocolitica and Y. pseudotuberculosis express variable O-antigen side chains, which have been theorized to enhance survival in the gastrointestinal tract (Reeves, P., Trends Microbiol., 3, 381-386, (1995)) and may inhibit complement-mediated lysis during invasive disease (Karlyshev, A. V., et al., Infect Immun, 69, 7810-7819, (2001)). In contrast, Y. pestis has a rough LPS phenotype with no O-specific side chains due to mutations in several O-antigen biosynthesis genes (Prior, J. G., et al., Microb. Pathog., 30, 48-57, (2001), Skurnik, M. P., A; Ervela, E, Mol Microbiol, 37, 316-330, (2000)).
Interestingly, genomic sequencing projects revealed that several virulence genes present in all three pathogenic Yersinia species have acquired mutations in Y. pestis that rendered them non-functional (Chain, P. S., et al., Proc. Natl. Acad. Sci. USA, 101, 13826-13831, (2004), Parkhill, J., et al., Nature, 413, 523-527, (2001)). Some of these encode invasin proteins that function during intestinal invasion in the enteropathogenic Y. enterocolitica and Y. pseudotuberculosis species, a host niche not colonized by Y. pestis (Simonet, M., et al., Infect Immun, 64, 375-379, (1996)). Other genes with lost function in Y. pestis include those involved in intermediary metabolism, and these functional losses are theorized to be part of the evolution of Y. pestis into an obligate parasitic species with the inability to survive outside the host (Parkhill, J., et al., Nature, 413, 523-527, (2001)). Research on the pathogenesis of Yersinia has largely been focused on the 70 kb virulence plasmid that is found in all pathogenic species of Yersinia. The sequence of this plasmid, called pYV in Y. pseudotuberculosis and pathogenic Y. enterocolitica and pCD1 in Y. pestis, is remarkably conserved between Y. pseudotuberculosis and Y. pestis (Chain, P. S., et al., Proc. Natl. Acad. Sci. USA, 101, 13826-13831, (2004)). Accordingly, the more distantly-related Y. enterocolitica species harbors a more divergent pYV plasmid, but the virulence gene sequences are highly conserved among all three species (Hu, P., et al., J Bacteriol, 180, 5192-5202, (1998), Snellings, N. J., et al., Infect Immun, 69, 4627-38, (2001)). Focus on this plasmid began when experiments determined that the pYV plasmid is absolutely required for virulence of Yersinia, although the plasmid alone cannot restore virulence to specific avirulent strains suggesting that non-pVY genes are also involved in pathogenesis (Heesemann, J., et al., Infect Immun, 46, 105-110, (1984), Heesemann, J. and R. Laufs, J Bacteriol, 155, 761-767, (1983)). A large locus on this plasmid encodes the Ysc-Yop system, a type III secretion system and its associated effector proteins. This system was the first example of a type III secretion apparatus, now identified in many animal and plant microbial pathogens (for review, see Cornelis, G. R., Nat. Rev. Mol. Cell. Biol., 3, 742-752, (2002)). The Yersinia Yop-Ysc secretion system includes “injectisome” proteins, translocator effector proteins, and Yop effector proteins. Electron microscopy and labeling studies with various type III secretory systems revealed that the injectisome proteins form a pore spanning the cytoplasmic and outer membranes of the bacteria and project a needle-like structure from the cell surface (Blocker, A., et al., Mol. Microbiol., 39, 652-663, (2001), Kimbrough, T. G. and S. I. Miller, Proc Natl Acad Sci USA, 97, 11008-11013, (2000), Kubori, T., et al., Science, 280, 602-605, (1998), Sukhan, A., et al., J Bacteriol, 183, 1159-1167, (2001)). The translocator proteins appear to interact with host macrophages and polymorphonuclear neutrophils (PMNs), forming a pore-like structure in the host cell membrane (Neyt, C. and G. R. Cornelis, Mol Microbiol, 33, 971-981, (1999)). The assembled secretion apparatus then allows the effector Yops to be translocated across the bacterial cell membranes and injected into the host cell, where they function by interfering with various immune response pathways (Bleves, S. and G. R. Cornelis, Microbes Infect., 2, 1451-1460, (2000), Cornelis, G. R., Nat. Rev. Mol. Cell. Biol., 3, 742-752, (2002)). The yadA gene is also present on the pYV plasmid, encoding the YadA adhesin with the ability to bind and adhere to eukaryotic cells (Eitel, J. and P. Dersch, Infect Immun, 70, 4880-91, (2002), Skurnik, M., et al., Infect Immun, 62, 1252-61, (1994)). This protein only appears to be functional in the enteropathogenic Yersinia, as a frameshift mutation in the Y. pestis yadA gene renders it non-functional (Hu, P., et al., J Bacteriol, 180, 5192-5202, (1998)).
The involvement of iron in Yersinia infections has long been established. For example, iron-overloaded patients such as those afflicted with β-thalassemia are highly susceptible to Yersinia infections (Farmakis, D., et al., Med. Sci. Monit., 9, RA19-22, (2003)). Furthermore, virulence could be restored in specific avirulent Y. pestis mutants by the addition of heme or heme-containing compounds (Burrows, T. W. and S. Jackson, Br. J. Exp. Pathol., 37, 577-583, (1956)). These early observations with Yersinia and other bacteria led researchers to study some of the microbial mechanisms of iron uptake. In mammalian hosts, available iron is extremely limited; intracellular iron is complexed with storage proteins, and extracellular iron is bound by the host proteins transferrin and lactoferrin. These iron-restricted conditions limit the growth of microbial invaders, thus acting as a defense barrier to infection. Many pathogens have evolved the ability to scavenge iron under these iron-poor conditions, effectively “stealing” iron from transferrin or heme-containing compounds. One of the most common mechanisms utilized by bacteria is the synthesis and secretion of siderophores, small molecules with a high affinity for iron (Andrews, S. C., et al., FEMS Microbiol. Rev., 27, 215-237, (2003)). The iron-siderophore complexes are bound by outer membrane receptors on the bacterial cell surface, and through the concerted action of outer membrane, periplasmic, and ABC transporter proteins, iron is transported into the cell. Other outer membrane receptors can directly bind heme and heme-containing compounds, scavenging the iron from these molecules. The role of several Yersinia iron uptake systems has been elucidated, while many more putative systems have been identified but not characterized.
Although Yersinia can use various siderophores produced by other bacteria and fungi to obtain iron, yersiniabactin is the only Yersinia-produced siderophore that has been detected (Baumler, A., et al., Zentralbl. Bakteriol., 278, 416-424, (1993), Rabsch, W. and G. Winkelmann, Biol Met, 4, 244-250, (1991), Reissbrodt, R. and W. Rabsch, Zentralbl Bakteriol Mikrobiol Hyg [A], 268, 306-317, (1988)). The yersiniabactin system is encoded by the ybt genes present on the chromosomal high-pathogenicity island (HPI), a locus that is associated with highly pathogenic strains of Yersinia (de Almeida, A. M., et al., Microb. Pathog., 14, 9-21, (1993), Rakin, A., et al., J Bacteriol, 177, 2292-2298, (1995)). The ybt genes encode proteins involved in the synthesis and secretion of the siderophore yersiniabactin (ybtS, irp1, irp2, ybtE, ybtT), as well as the cytoplasmic (ybtP, ybtQ) and outer membrane proteins (psn/fyuA) required for uptake of the iron-yersiniabactin complexes (Carniel, E., Microbes Infect., 3, 561-569, (2001)). Mutations in genes for yersiniabactin synthesis and/or uptake resulted in reduced Yersinia virulence in mouse models of infection (Bearden, S. W., et al., Infect. Immun., 65, 1659-1668, (1997), Brem, D., et al., Microbiology, 147, 1115-1127, (2001), Rakin, A., et al., Mol Microbiol, 13, 253-263, (1994)), indicating that this system is an important virulence factor in Yersinia pathogenesis. The nucleotide sequence of the ybt genes are at least 97% identical between the three pathogenic Yersinia species (Carniel, E., Microbes Infect., 3, 561-569, (2001), Chain, P. S., et al., Proc. Natl. Acad. Sci. USA, 101, 13826-13831, (2004)), and the Y. pestis and Y. pseudotuberculosis ybt systems were demonstrated to be interchangeable (Perry, R. D., et al., Microbiology, 145 (Pt 5), 1181-1190, (1999)). These analyses indicated that the functions of these homologs are likely conserved among the three species. Furthermore, the HP1 has been discovered in various pathogenic species including some strains of E. coli, Citrobacter, and Klebsiella (Bach, S., et al., FEMS Microbiol. Lett., 183, 289-294, (2000)). The Ybt proteins expressed by these organisms are quite similar; indeed, antibodies raised against several of the Yersinia Ybt proteins recognized the corresponding proteins from the other pathogens (Bach, S., et al., FEMS Microbiol. Lett., 183, 289-294, (2000), Karch, H., et al., Infect Immun, 67, 5994-6001, (1999)). These results suggest that the acquisition of the ybt system is relatively recent among these pathogens and may have contributed to the invasive phenotypes associated with many of these serotypes.
Several additional ybt-independent iron uptake systems have been detected in Yersinia species based on mutation analysis, homology to known iron acquisition proteins, or the presence of iron-responsive regulatory elements. One such regulatory element is the “Fur box,” a nucleotide sequence that binds the regulatory protein Fur when it is complexed with iron. The binding of Fe-Fur to a Fur box represses transcription of downstream promoters, and when iron becomes limiting, apo-Fur dissociates from DNA and transcription is derepressed. Fur and its homologs have been found in most species of bacteria, and regulate many genes in addition to iron uptake systems in diverse organisms (Campoy, S., et al., Microbiology, 148, 1039-1048, (2002), Horsburgh, M. J., et al., Infect Immun, 69, 3744-3754, (2001), Sebastian, S., et al., J Bacteriol, 184, 3965-3974, (2002), Stojiljkovic, I., et al., J Mol Biol, 236, 531-545, (1994)). Analysis of the Y. pestis genome identified many genes with Fur boxes upstream of their respective promoters, most of which encoded proteins with homology to known iron uptake systems (Panina, E. M., et al., Nucleic Acids Res, 29, 5195-5206, (2001)). Although few of these genes have been studied for function, several appear to encode iron-siderophore receptor proteins (omrA, irgA, itrA, ihaB, fauA) and iron ABC transporters (itsTUS, itpPTS). Since Yersinia can utilize siderophores produced by other organisms, these proteins may be responsible for the “siderophore piracy” observed with Yersinia. Such methods of iron acquisition are common among bacterial pathogens.
Several studies have elucidated the functions of other putative iron uptake systems. For example, the Hmu system of Y. pestis was demonstrated to acquire iron through the uptake of heme and heme-containing compounds (Hornung, J. M., et al., Mol Microbiol, 20, 725-39, (1996)). Although the ability to use heme as an iron source seems advantageous for a pathogen, the Y. pestis hmu mutant was fully virulent in a mouse model of infection (Thompson, J. M., et al., Infect Immun, 67, 3879-92, (1999)). A second putative heme-uptake system was identified in Y. pestis on the basis of sequence homology. The has genes of Y. pestis are homologs of the hemophore-dependent heme acquisition genes of Pseudomonas and Serratia (Rossi, M. S., et al., Infect Immun, 69, 6707-6717, (2001)). In these organisms, a hemophore (HasA) is secreted that binds heme and delivers it to bacterial surface receptors (HasR) to transport heme into the cell. The Y. pestis HasA protein was determined to be Fur-regulated, secreted, and capable of binding heme. However, a mutation in these genes had no effect on virulence in the mouse, even when a double mutant was tested (Rossi, M. S., et al., Infect Immun, 69, 6707-6717, (2001)). Therefore, the roles of the putative heme uptake systems in pathogenesis remain elusive, and may indicate that heme uptake is more important during infection of non-murine hosts.
The functions of two putative iron ABC transport systems have also been studied in Yersinia. The Yfe system can transport iron and manganese in Y. pestis, and yfe mutants demonstrated reduced virulence in mouse models of infection (Bearden, S. W. and R. D. Perry, Mol. Microbiol., 32, 403-414, (1999)). The second putative iron ABC transporter proteins are encoded by the yfu genes, identified by the presence of an upstream Fur box (Gong, S., et al., Infect. Immun., 69, 2829-2837, (2001)). When expressed in E. coli, the yfu genes restored growth in iron-poor media; however, comparable studies in Y. pestis failed to determine a role for Yfu in iron acquisition, and the yfu-mutant showed no defect in mouse virulence (Gong, S., et al., Infect. Immun., 69, 2829-2837, (2001)).